用非线性模型估测恒温和变温下棉铃虫蛹的发育率

为了深入分析和探讨昆虫发育与环境温度的关系, 在恒温（15~37℃）和交替变温（12/18~34/40℃）下测定了棉铃虫Helicoverpa armigera蛹的发育历期(d),分别用线性模型和非线性模型（Logan模型﹑Lactin模型和王氏模型）拟合其发育率(1/d)数据。结果表明,这3个非线性模型能更准确地描述发育率与温度之间的曲线关系,判定系数（R2）在0.9878～0.9991之间。对全部观测数据的进一步研究表明,只要有6个分布合适的观测数据,就可以用这些非线性模型获得相当满意的估测效果。如果缺乏高温下的测定数据,用非线性模型预测的昆虫发育率可能失真。分析了蛹在恒温和变温下发育率差异的可能原因,讨论了应用这3个非线性模型预测蛹期发育的优点和缺点,指出用非线性模型取代线性日·度模型进行害虫发生预测和益虫饲养管理的合理性和必要性。     

基于DWT-GA-PLS的土壤碱解氮含量高光谱估测方法

以山东齐河县为研究区,实地采集土壤样本,在土样高光谱测试并进行一阶导数变换的基础上,先运用离散小波变换(DWT)对土壤光谱去噪降维,然后采用遗传算法(GA)筛选土壤碱解氮定量估测模型的参与变量,最后应用偏最小二乘(PLS)回归构建土壤碱解氮含量的估测模型.结果表明: 离散小波变换结合遗传算法和偏最小二乘法(DWT-GA-PLS)用于土壤碱解氮含量定量估测,不仅可压缩光谱变量、减少模型参与变量,而且可改善模型估测准确度;较之于采用土壤全谱,小波离散分解1～2层低频系数构建的模型在参与变量大幅减少的情况下,取得更准确或与之相当的预测结果,其中,基于第2层小波低频系数采用GA筛选变量构建的PLS模型的预测效果表现最好,预测R²达到0.85,RMSE为8.11 mg·kg^-1,RPD为2.53.说明DWT-GA-PLS用于土壤碱解氮含量高光谱定量估测的有效性.     

基于小波变换的土壤有机质含量高光谱估测

利用统计分析方法选取了土壤N、P、K元素含量近似而有机质含量差异较大的样本60个,通过高光谱探测分析获得样本反射率对数的一阶导数光谱,采用Bior 1.3函数进行多层离散小波分解,剔除低频近似信号和高频噪声信号,得到反映土壤理化参数的特征光谱曲线;采用相关分析筛选土壤有机质含量的显著相关波段,基于显著相关波段和特征光谱曲线分别构建土壤有机质含量高光谱多元回归估测模型;通过比较分析,确定了提取土壤有机质特征光谱的最佳小波分解尺度并构建了最佳预测模型.结果表明： 提取土壤有机质特征光谱的最佳小波分解层数是9层,其次是8层和10层;基于小波9层分解特征光谱曲线的有机质含量估测模型最佳,其决定系数（R²）为0.89,比基于显著相关波段构建模型的R²增加了0.31,比基于原始光谱所构建模型的R²增加了0.10.     

近红外光谱分析法测定东北黑土有机碳和全氮含量

以我国东北黑土为研究对象,分析了2004-2005年采集的136个土壤样品在3699～12000 cm^-1范围的近红外光谱,利用偏最小二乘法建立了原始光谱吸光度与土壤有机碳、全氮和碳氮比之间的定量分析模型.结果表明：土壤有机碳和全氮的模型拟合效果良好,决定系数R²分别为0.92和0.91（P<0.001）,相对分析误差RPD分别为3.45和3.36,利用该模型对验证样本土壤有机碳和全氮的预测值与实测值之间的相关系数分别为0.94和0.93（P<0.001）,表明可以用近红外光谱分析法对黑土有机碳和全氮含量进行测定.但是利用近红外光谱分析法对土壤碳氮比的预测并不理想,虽然验证样本集黑土碳氮比模型预测值与实测值呈显著相关（r＝0.74,P<0.001）,但是校正模型的R²为0.61,RPD仅为1.61,建立的模型不能对黑土碳氮比做出合理的估测.     

基于无人机的冬小麦拔节期表层土壤有机质含量遥感反演

快速监测大面积分布的盐渍化麦田土壤有机质含量,可为推进盐渍土改良和促进碳循环研究提供数据支撑。通过野外采样与获取无人机遥感影像,分别基于裸土和植被情况,采用多元线性回归(MLR)、偏最小二乘回归(PLSR)和支持向量机回归(SVR)3种方法,建立区域有机质含量遥感模型,并进行检验和对比,确定最优的土壤有机质含量反演模型;最后基于最优模型进行研究区表层土壤有机质的反演,并与插值结果进行比较。结果表明: 经5×5的中值滤波处理后的光谱与土壤表层有机质对应最优;3种模型中,SVR模型的预测精度最高,PLSR次之,MLR效果最差。对比两种变量的建模效果,基于植被的SVR建模效果最好,其建模决定系数(R²)、均方根误差(RMSE)分别为0.89、0.20,验证R²、RMSE分别为0.82、0.24;基于裸土的建模效果不理想,最优的也是SVR模型,其建模R²、RMSE分别为0.63、0.26,验证R²、RMSE分别为0.61、0.25。根据最优模型反演得到该区域有机质含量为17.51～22.53 g·kg^-1,平均值为19.51 g·kg^-1,与实地调查结果较为一致;插值结果与反演结果相比,精度受到限制。综上,基于无人机多光谱可以对盐渍土冬小麦拔节期土壤有机质含量进行快速、大范围精准估测。     

新疆云杉蓄积与地上生物量模型拟合分析

目的：建立新疆云杉蓄积及地上生物量模型,为在林分尺度上估算云杉碳储量及生产力提供基础数据。方法：利用收获法采集西伯利亚云杉(Picea obovata)与雪岭云杉(Picea schrenkiana)各50株,以蓄积为自变量,地上生物量为因变量,采用4种生物量模型进行回归分析构建模型,综合模型评价、参数估计值的稳定性及相对误差筛选出最优生物量模型。结果：确定W=0.515V^0.926 (R²=0.980)为西伯利亚云杉估测模型,W=0.541V^0.953 (R²=0.954)为雪岭云杉估测模型。结论：新疆云杉蓄积和地上生物量极显著相关,4种估测模型以幂函数拟合效果最优。     

桃小食心虫在不同温度下的实验种群生命表

为探索温度对桃小食心虫Carposina sasakii生长发育和繁殖的影响,在室内17, 20, 23, 26, 29和32(±1) ℃,80%±7% RH和15L: 9D条件下,测定了桃小食心虫各发育阶段的历期、存活率和/或产卵量, 组建了桃小食心虫的实验种群生命表.结果表明, 桃小食心虫各虫态的发育历期随温度的升高而缩短; 初孵幼虫的蛀果率随温度升高而提高,幼虫脱果率与温度之间呈抛物线关系：y=-0.5638x²+27.882x-269.18 (R²=0.9801,P＜0.01);结茧率和羽化率在17～29℃间无明显变化,但在32℃时则明显降低;雌蛾的产卵量和寿命随温度的升高而降低,23℃时雌蛾交配率最高.生命表分析表明,种群趋势指数在17～29℃间均大于1,26℃时内禀增长率最高;世代存活率与温度的关系可用S=-0.073x³+4.626x²-92.019x+596.57（R²=0.9832）表示; 内禀增长率与温度之间的关系可用r_m=-0.0008x²+0.0409x-0.4438（R²=0.9851）描述。 据此得出,23～26℃是最适宜桃小食心虫生长发育和繁殖的温度范围。     

基于相关分析和偏最小二乘回归的黄绵土土壤全氮和碱解氮含量的高光谱预测

以采取植被恢复措施的陕西省吴起县为研究区,实地采集24个土壤剖面不同层次的黄绵土土样100个,在进行土壤样本全氮（TN）和碱解氮（AHN）含量及实验室反射光谱数据测量和分析的基础上,用相关分析（CA）结合偏最小二乘回归（PLS）方法建立黄绵土土壤TN和AHN含量的校正模型,并用独立样本对校正模型进行验证.结果表明： 利用6种光谱变换方式建立的校正模型中,微分光谱建立的校正模型是预测研究区土壤TN含量的最佳模型,校正和验证R²分别为0.929和0.935,均方根误差（RMSE）分别为0.045和0.047 g·kg^-1,相对预测偏差（RPD）为3.12;而归一化变换建立的校正模型是预测土壤AHN含量的最佳模型,校正和验证R²分别为0.873和0.773,RMSE分别为9.946和16.204 mg·kg^-1,RPD为1.538.所建立的全氮预测模型可以对0～40 cm土层的TN进行有效预测,而碱解氮的预测模型对同一深度只能进行粗略预测.本研究为采取植被恢复措施的退化生态系统区黄绵土土壤全氮的快速预测提供了一种较好的方法,但是对于碱解氮的准确、快速预测,需要进一步研究.     

普洱季风常绿阔叶林乔木物种多样性高分辨率遥感估测

季风常绿阔叶林是我国南亚热带典型的地带性植被,也是云南省普洱地区重要森林类型。季风常绿阔叶林乔木物种多样性遥感估测对研究区域尺度生物多样性格局及其规律具有重要作用。根据光谱异质性假说和环境异质性假说,首先使用1m空间分辨率的机载高光谱数据和激光雷达数据提取了光谱多样性特征和垂直结构特征。然后利用基于随机森林算法的递归特征消除方法选择对研究区森林乔木物种多样性指数具有较好解释能力的遥感特征,并对Shannon-Winner物种多样性指数进行建模、制图。研究结果表明：(1)基于机载LiDAR数据提取的垂直结构特征和机载高光谱数据提取的光谱多样性特征均对研究区森林乔木物种多样性具有较好的解释能力,随机森林模型估测结果分别为R²=0.48,RMSE=0.46和R²=0.5,RMSE=0.45;两种数据源融合可以进一步提高遥感数据的森林乔木物种多样性估测精度,随机森林估测模型R²和RMSE分别为0.69和0.37。(2)机载激光雷达数据对研究区针阔混交林乔木物种多样性的估测能力优于机载高光谱数据。(3)机器学习方法有助于从高维遥感...     

三种森林生物量估测模型的比较分析

森林生物量的定量估算为全球碳储量、碳循环研究提供了重要的参考依据。该研究采用黑龙江长白山地区的TM影像和133块森林资源一类清查样地的数据, 选取地学参数、遥感反演参数等71个自变量分别构建多元逐步回归模型、传统BP (back propagation)神经网络模型和基于高斯误差函数的BP神经网络改进模型(Gaussian error function, Erf-BP), 进而估算该地区的森林生物量, 并进行比较分析。结果表明, 多元逐步回归模型估测的森林生物量预测精度为75%, 均方根误差为26.87 t·m^-2; 传统BP神经网络模型估测森林生物量的预测精度为80.92%, 均方根误差为21.44 t·m^-2; Erf-BP估测森林生物量的预测精度为82.22%, 均方根误差为20.83 t·m^-2。可见, 改进后的Erf-BP能更好地模拟生物量与各个因子之间的关系, 估算精度更高。     

Temperature‐dependent developmental model of Neoseiulus californicus (McGregor) (Acari,Phytoseiidae)

The developmental time and survival of immature stages of Neoseiulus californicus were studied at nine constant temperatures (12, 16, 24, 24, 28 32, 36, 38 and 40°C), 60–70% RH, and a photoperiod of 16 : 8 (L : D) h. The total mortality of immature N. californicus was lowest at 24°C (4.5%) and highest at 38°C (15.2%). The total developmental time decreased with increasing temperature between 12°C (18.38 days) and 32°C (2.98 days), and increased beyond 32°C. The relationship between the developmental rate and temperature was fitted by five nonlinear developmental rate models (Logan 6, Lactin 1, 2 and Briere 1, 2). The nonlinear shape of temperature development was best described by the Lactin 1 model (r² = 0.98). The developmental variation of each stage was well described by the three‐parameter Weibull distribution model (r² = 0.91–0.93). The temperature‐dependent developmental models of N. californicus developed in this study could be used to determine optimal temperature conditions for its mass rearing, to predict its seasonal population dynamics in fruit tree orchards or greenhouse crops, or to develop a population dynamics model of N. californicus.     

USING TREND INDICES FOR ENDANGERED SPECIES

Initial appraisals of the status of endangered large-mammal populations may have to depend on indices of population trend. Such indices may possibly be improved by using auxiliary variables. Various models were studied for populations of the Florida manatee (Trichechus manatus latirostris), Yellowstone grizzly bear (Ursus arctos horribilis), and Hawaiian monk seal (Monachus schauinslandi). Several criteria for checking validity of the fitted models were considered, and the simple R² criterion appears to provide useful comparisons. Multiple regression models overestimated the rate of change of the East Coast manatee population as determined from three other sources (a covariance model, a non-linear model, and the rate estimated from reproductive and survival data). A multiple regression model for grizzly bears using three auxiliary variables exhibited a fairly high R² (0.84) and appeared to provide a better fit than did a non-linear model. A beach count index for Hawaiian monk seals seemed to be unreliable for year-to-year comparisons in contrast to total population counts and estimates from a capture-recapture method. The use of auxiliary variables for checking and improving trend index data appears feasible and well worthwhile.     

QSAR studies on 4-anilino-3-quinolinecarbonitriles as Src kinase inhibitors using robust PCA and both linear and nonlinear models

Quantitative structure-activity relationship (QSAR) studies have been carried out on 4-anilino-3-quinolinecarbonitriles, a set of novel Src kinase inhibitors, with the aim of dissecting the structural requirements for Src inhibitory activities. After outlier identification using robust principal component analysis (robust PCA), linear models based on forward selection combined with multiple linear regression, (FS-MLR), enhanced replacement method followed by multiple linear regression (ERM) and a nonlinear model using support vector regression (SVR) were constructed and compared. All models were rigorously validated using leave-one-out cross-validation (LOOCV), 5-fold cross-validations (5-CV) and shuffling external validation (SEVs). ERM seems to outperform both FS-MLR and SVR evidenced by better prediction performance (n?=?35, R²_training?=?0.918, R²_pred?=?0.928). Robustness and predictive ability of ERM model were also evaluated. The generated QASR model revealed that the Src inhibitory activity of 4-anilino-3-quinolinecarbonitriles could be associated with the size of substituents in the C7 position and the steric hindrance effect. The results of the present study may be of great help in designing novel 4-anilino-3-quinolinecarbonitriles with more potent Src kinase inhibitory activity.     

Effect of temperature on reproductive parameters and longevity of <Emphasis Type="Italic">Tyrophagus putrescentiae</Emphasis> (Acari: Acaridae)

The effect of temperature on reproductive parameters and longevity of the mold mite, Tyrophagus putrescentiae (Schrank) was examined at seven constant temperatures, ranging from 10 to 34 °C, and a relative humidity of 90±5%. Preoviposition period and fecundity were adversely affected by extreme temperatures and the oviposition period increased as temperature was reduced. Different patterns were observed for longevity data for males and females, with greater longevities for males at intermediate temperatures and more similar values for both sexes at extreme temperatures. Polynomial and non-linear models provided a good fit of the relationship of reproductive and longevity parameters with temperature. The effect of temperature on the intrinsic rate of increase of T. putrescentiae populations was established by the non-linear Lactin model. The optimum temperature for development was obtained at 30 °C. At this temperature, the population doubling time is 1.75 days. The lower and upper thresholds for T. putrescentiae populations were established at 10.4 and 34.8 °C, respectively. Altogether, these data provide basic information to develop sound physical control strategies of the mold mite.     

Estimating temperature‐dependent developmental rates of potato tuberworm,Phthorimaea operculella (Lepidoptera: Gelechiidae)

Abstract The potato tuberworm, Phthorimaea operculella (Zeller) (Lepidoptera: Gelechiidae), is the most destructive pest of potato, Solanum tuberosum L. (Solanaceae), in tropical and subtropical regions in both field and storeroom situations. The modeling of temperature‐dependent development can be useful in forecasting occurrence and population dynamics of the pests. Published developmental parameters for this pest vary greatly for many reasons. We determined temperature‐dependent development of P. operculella at seven constant temperatures (16, 20, 24, 28, 32, 34 and 36 °C). Developmental period of whole immature stage (egg to the end of the pupal stage) varied from 75.5 days at 16 °C to 17 days at 32 °C. The population failed to survive at 36 °C. The observed data was modeled to determine mathematical functions for simulating P. operculella development in each stage of development and overall. Two linear models, ordinary linear regression and the Ikemoto linear model were used to describe the relationship between temperature and development rate of the different stages of P. operculella and estimating the thermal constant and lower temperature threshold. The lower temperature threshold (t) and thermal constant (k) of whole immature stage were estimated to be 11.6 °C and 338.5 DD by Ikemoto linear model, and the estimated parameters were not substantially different with those estimated by ordinary linear models. Different models provided a better fit to the various developmental stages. Of the eleven nonlinear models fitted, the Beriere‐1, Logan‐6 and Lactin‐1 model was found to be the best for modeling development rate of egg, larva and pupa of P. operculella, respectively. Phenological models based on these findings can be part of a decision‐support tool to improve the efficiency of pest management programs.     

星载大光斑LiDAR与HJ-1A高光谱数据联合估测区域森林地上生物量

以吉林省汪清林业局经营区为研究区,利用HJ-1A/HSI高光谱数据和ICESat-GLAS波形数据,估测区域森林地上生物量。从平滑后的GLAS波形数据中提取波形长度W和地形坡度参数TS,建立GLAS森林最大树高估测模型;从GLAS波形数据中提取能量参数I(植被回波能量Ev和回波总能量E之比),建立GLAS森林郁闭度估测模型;利用GLAS估测的森林最大树高和森林郁闭度联合建立森林地上生物量模型。由于GLAS呈离散条带状分布,无法实现区域估测,因此研究将GLAS波形数据与HJ-1A/HSI高光谱数据联合,基于支持向量回归机算法实现森林地上生物量区域估测,得到研究区森林地上生物量分布图。研究结果显示,基于W和TS建立的GLAS森林最大树高估测模型的adj.R~2=0.78,RMSE=2.51m,模型验证的adj.R~2=0.85,RMSE=1.67m。地形坡度参数TS能够有效的降低地形坡度的影响;当林下植被高度为2m时,得到的基于参数I建立的GLAS森林郁闭度估测模型效果最好,模型的adj.R~2=0.64,RMSE=0.13,模型验证的adj.R~2=0.65,RMSE=0.12。利用森林最大树高和森林郁闭度建立的森林地上生物量模型的adj.R~2=0.62,RMSE=10.88 t/hm~2,模型验证的adj.R~2=0.60,RMSE=11.52 t/hm~2。基于支持向量回归机算法,利用HJ-1A/HSI和GLAS数据建立的森林地上生物量SVR模型,生成了森林地上生物量分布图,利用野外数据对得到的分布图进行验证,验证结果显示森林地上生物量估测值与实测值存在很强的线性关系(adj.R~2=0.62,RMSE=11.11 t/hm~2),能够满足林业应用的需要。因此联合ICESat-GLAS波形数据与HJ-1A高光谱数据,能够提高区域森林地上生物量的估测精度。     

Improved age determination of blood and teeth samples using a selected set of DNA methylation markers

Age estimation from DNA methylation markers has seen an exponential growth of interest, not in the least from forensic scientists. The current published assays, however, can still be improved by lowering the number of markers in the assay and by providing more accurate models to predict chronological age. From the published literature we selected 4 age-associated genes (ASPA, PDE4C, ELOVL2, and EDARADD) and determined CpG methylation levels from 206 blood samples of both deceased and living individuals (age range: 0–91 years). This data was subsequently used to compare prediction accuracy with both linear and non-linear regression models. A quadratic regression model in which the methylation levels of ELOVL2 were squared showed the highest accuracy with a Mean Absolute Deviation (MAD) between chronological age and predicted age of 3.75 years and an adjusted R² of 0.95. No difference in accuracy was observed for samples obtained either from living and deceased individuals or between the 2 genders. In addition, 29 teeth from different individuals (age range: 19–70 years) were analyzed using the same set of markers resulting in a MAD of 4.86 years and an adjusted R² of 0.74. Cross validation of the results obtained from blood samples demonstrated the robustness and reproducibility of the assay. In conclusion, the set of 4 CpG DNA methylation markers is capable of producing highly accurate age predictions for blood samples from deceased and living individuals     

Soil Temperature, Not Aboveground Plant Productivity, Best Predicts Intra-Annual Variations of Soil Respiration in Central Iowa Grasslands

Soil respiration (R_SOIL) is the second largest carbon flux between terrestrial systems and the atmosphere, with a magnitude 10 times greater than anthropogenic carbon dioxide production. Therefore, it is important that we understand, and be able to predict, how R_SOIL responds to climate change. Although a positive, significant temperature effect on R_SOIL has long been recognized, recent studies emphasize the overriding importance of current photosynthesis in controlling R_SOIL. We tested the hypothesis that model inclusion of intra-annual variations in aboveground net primary productivity (ANPP) significantly improves R_SOIL estimates over predictions based on soil temperature alone. We also evaluated the possibility that canopy production is less directly linked to R_SOIL, by testing the hypothesis that intersite differences in R_SOIL correlate more strongly with root biomass than with ANPP. We tested these hypotheses by measuring R_SOIL, ANPP, and root biomass at four Iowa grasslands that differed in aboveground growth phenology and productivity. Among all sites, intra-annual variations in R_SOIL were most strongly related to soil temperature (R ² = 0.89), not ANPP (R ² = 0.53). All sites responded identically to changes in soil temperature (site-by-temperature P = 0.53), but inconsistently to variation in aboveground dynamics (site-by-canopy P < 0.0001). Incorporating canopy dynamics into temperature-based predictive models improved model R ² by a maximum of 0.01. Among-site differences in R_SOIL were related to root biomass (P < 0.001) but not ANPP (P = 0.34). We found no useful linkage between canopy characteristics and intra-annual or site-specific R_SOIL predictions, perhaps because shoot and root dynamics were not consistently linked through time or among sites.